Hermes Agent Integration Guide

Target audience: Developers, Admin Users, Member Users Last updated: 2026-08-06

Hermes Agent officially provides a custom OpenAI Compatible Endpoint wizard. Using the wizard is more reliable than manually assembling configuration commands and keeps up with configuration structure changes across versions.

Installation

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

After installation, run in the terminal:

hermes model

Select Custom endpoint and fill in the following:

API Base URL: https://<your-api-endpoint>
API Key: YOUR_API_KEY
Model: deepseek-v4-flash

For the API Base URL, enter the root address. The client will append /v1/chat/completions automatically.

Use the endpoint address shown on the "API Key" page in the console.

When to Use Manual Configuration

The wizard writes the Provider, model, and address into ~/.hermes/config.yaml. You only need to edit this file manually for automated deployments or when troubleshooting migration issues. Use the file generated by the current version as your template for field structure.

Model fields in the legacy .env file may no longer take effect. If the connection still points to the old Provider after configuration, run hermes model to reselect, and check whether config.yaml retains an old provider or base_url.

Verification

Start Hermes, conduct a round of basic Q&A, then test a read-only tool. Use /model within a session to switch between already-configured models. To add or modify a Provider, return to the terminal and run hermes model.